The study split an educational attainment genetic score into an intelligence component and a behavioural one, and tested each against Alzheimer's outcomes in 3,483 participants. The intelligence component was inversely associated with diagnosis in all four cohorts, and with post-mortem tangle stage in the two brain-donation cohorts. The behavioural component showed no association with diagnosis, blood biomarkers, amyloid scan status or post-mortem plaque scores.
Why it is interesting: It separates the protective association of education into a part that tracks pathology and a part that does not.
Abstract — the authors' own words, CC BY
Objective
Some individuals avoid Alzheimer's disease (AD) pathology as they age, or retain cognition despite substantial pathology, suggesting mechanisms of resistance or resilience to neurodegeneration. Educational attainment (EA) is associated with reduced risk of cognitive decline, although the underlying mechanisms remain unclear. We investigated whether genome-wide polygenic scores (GPS) based on cognitive and non-cognitive components of EA, reflecting intelligence and behavioral/social traits, respectively, show differential associations with AD diagnosis and neuropathology. Methods
Associations between GPS for cognitive and non-cognitive components of EA and AD outcomes were examined across four cohorts: two cohorts included living participants with biomarker assessments (BioHermes-1, n = 770; Alzheimer's Disease Neuroimaging Initiative [ADNI], n = 1,361), and two included postmortem neuropathological evaluations (Religious Orders Study and Rush Memory and Aging Project [ROSMAP], n = 841; Brain for Dementia Research [BDR], n = 511). Outcomes included clinical diagnosis, amyloid status, plasma neurodegenerative biomarkers, and neuropathological measurements. Results
We found a significant negative association between cognitive EA GPS and clinical diagnosis of AD across all four cohorts and with Braak stage in ROSMAP and BDR cohorts. In contrast, the GPS for non-cognitive component of EA showed no significant association with AD diagnosis, neurodegenerative plasma biomarkers or amyloid positron emission tomography (PET) status and the Consortium to Establish a Registry for Alzheimer's Disease (CERAD) scores. Interpretation
These results suggest a brain maintenance mechanism that supports the brain's ability to resist to changes in neuronal integrity. This indicates that interventions focused on increasing EA solely by additional years of schooling are unlikely to affect AD incidence, whereas alternative approaches targeting the cognitive component of EA (intelligence) may offer greater potential for prevention. ANN NEUROL 2026.
Cite this item
BibTeX
Free sign-up. A digest of the new papers twice a week, grouped by domain.